[mlpack] GSOC 2016 Contribute to MLPACK

[Ryan Curtin]

On Tue, Mar 08, 2016 at 11:37:40AM +0530, Divyam Khandelwal wrote:
> Hello Sir Ryan Curtin and Sumedh Ghaisas,
>    I read  about archives in mailing list so I firstly decided to
> build mlpack library and I had compile mlpack tutorial programs and
> now I am comfortable with it.Thanks for helping in completing my first
> step. I am Third Year B.Tech student CS Undergraduate from Vishwakarma
> Institute of Technology,Pune.I am an active C,C++ coder for many years
> now. I am currently studying Recommender Systems and I want to explore
> it lot by learning new optimization methods and by implementing
> it.However, while it is possible for me to improve my knowledge of
> recommender system  on the side while studying machine learning (most
> of my recommender system was self-taught anyway).I’ve always had a
> passion for both programming and mathematics.
>    So I want to move towards the second step to solve bug please
> direct me how to proceed toward it
> 
>    I'm interesrted in Alternatives to neighborhood-based collaborative filtering
>    I read about the paper attached on the site. Even I read the matrix
> factorization method of Amazon and Netflix which they used and
> understood it.
>    The reference material I read uptill now is
>    =>The given paper
>    =>Slideshare giving a brief idea about matrix factorization
>    https://www.google.co.in/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&cad=rja&uact=8&ved=0ahUKEwjspY2ngK_LAhXOTI4KHU5NC8kQFgg8MAQ&url=http%3A%2F%2Fwww.slideshare.net%2Fstudentalei%2Fmatrix-factorization-techniques-for-recommender-systems&usg=AFQjCNEtwt1uY6UqtIRV595sg9KSkCX2vg&sig2=OWdhVM1sjWXiSRw10VaCGA&bvm=bv.116274245,d.c2E
>    =>Matrix Factorization Techniques for Recommender Systems
>  By Yehuda Koren, Yahoo Research
>        Robert Bell and Chris Volinsky, AT&T Labs—Research
> 
> Please recommend what literature should i read and what type of bugs
> should i solve in this topic
>      Thanking You

Hi Divyam,

There's no need to send your email to the mailing list twice---sometimes
it just takes a little while for someone to have enough free time to
respond.

Here is some more information on the CF project:

https://mailman.cc.gatech.edu/pipermail/mlpack/2016-March/000807.html

It would probably be a good idea to familiarize yourself with the papers
corresponding to algorithms that have already been implemented; take a
look through the comments in src/mlpack/methods/cf/ and
src/mlpack/methods/amf/ and you should be able to find papers to read
(usually they are listed in BiBTeX format, so you can probably grep for
the words 'title = ' or 'title=' to find them).

I don't know if there are currently any open bugs that correspond to the
CF code, but you can search on the Github issues tracker:

https://github.com/mlpack/mlpack/issues/

You could also look through the existing code and see if there are any
bugs you can find or improvements that you can think of.

I hope that this is helpful; let me know if I can clarify anything.

Thanks,

Ryan

-- 
Ryan Curtin    | "Moo."
ryan at ratml.org |   - Eugene Belford